By Jerri-Lynn Scofield, who has worked as a securities lawyer and a derivatives trader. She is currently writing a book about textile artisans.
One depressing aspect of the COVID-19 pandemic is to see two of the usual suspects – Big Tech and Big Pharma – trying to distort what we hear, learn, and understand about the ongoing situation. Each has a dog in the COVID-19 fight and seeks to convince us they have viable approaches that just happen to maximize their profits, whether it be the necessity of proprietary apps or the need to rely on new proprietary approaches – such as a new vaccine and expensive drugs under patent that have yet to be put forward,
I originally intended to combine discussion of these two grifts into one post. But in the interest of keeping the post short and manageable and also in concentrating discussion in comments on what are actually two separate problems – albeit each falling under the general heading of profiteering – I have decided to split the post, and discuss Big Tech today, and Big Pharma tomorrow.
Contact Tracing
I have written extensively about Hong Kong’s success in fighting COVID-19. The city’s effective coronavirus control strategy relies on tried and tested public health measures to control spread of an infectious disease: the use of masks; social distancing; and testing and tracing. Background assumptions I should mention are that the city has excellent health care, and public health officials have ample experience and enjoy public trust,
They didn’t hit on their COVID-19 strategy by serendipity. Previous experience in 1968 with what’s still known as the Hong Kong’ flu and SARS in 2003 informed their infectious disease control strategy.
Beginning in March, I have written at least five previous posts discussing aspects of Hong Kong’s approach,comparing the city’s response and record to that of New York, as they are similarly sized and densely populated (see for my most recent take, No, We Don’t Need to Place Our Faith in Downloading Some Untested, Privacy-Infringing App as the Only Possible COVID-19 Slayer; Why Don’t We Look to Places that Have Successfully Limited Disease Spread and Copy Their Policies?)
Now, to be sure, Hong Kong has recently seen an uptick in cases: today the South China Morning Post reported the total number of cases now stands at 1569 and 8 people have died. And the public health system has reacted quickly and hard to get control again over the spread of this very sneaky disease that exploits any weakness or lack of vigilance.
Conflating Contact Tracing and Contact Tracing By App
An integral part of Hon Kong’s successful strategy is effective contact tracing. Yet this is not the same thing as contact tracing via app. In a recent article in The Conversation, Digital contact tracing’s mixed record abroad spells trouble for US efforts to rein in COVID-19, Bhaskar Chakravorti, dean of global business at Tufts University’s Fletcher School conflated contact tracing with digital contact tracing, aka contact tracing via app.
Ignacio, in comments on a previous post, noted that contact tracing via app alone isn’t very good contact tracing. In fact, if I understand him correctly, it can be worse than useless. Maybe, if he reads this, he’ll pipe up in comments with a more thorough explanation, with some details as to why this is so.
It merely alerts you that you at some point were near to someone who has subsequently tested positive for the disease. But, what good does that do? Especially with the increasing number of infections.In fact, it may also underestimate your true chance of infection, since I understand these apps don’t necessarily work very well indoors, and we’re coming to learn that there, particularly in poorly ventilated places, you are at great if not greatest risk of contracting the virus.
Instead, effective contact tracing requires more than a determination that you may have been vicinity of someone who subsequently tested positive for the disease. In fact, contact tracing is not simple, as Dr.Sarah Borwein pointed out to me in a discussion I included in a previous post, Contact Tracing Via Old Shoe-Leather Epidemiology While Spurning the Techno-Fix Fairy: How Hong Kong Quells COVID-19 Without Killing Civil Liberties:
[Crucial to this success] is the importance of painstaking test and trace measures to plot and thwart the course of COVID-19 spread.
Sarah described what has been undertaken as “old shoe-leather epidemiology.” No app. No technofix fairy. Just hard work.
I asked her to explain what test and trace means to Hong Kong health authorities.
Sarah Borwein: So we had 21 days with no local cases and then a case was detected 2 days ago, and now her grand-daughter and husband have tested positive. What they are doing reflects their strategy:
They did extensive interviews with the index case (the 66 year old grandmother) and retraced everywhere she’d been in the 2-3 days prior to getting sick – every market stall etc. She looks after her 5 year old grand-daughter who is also positive – so they have also traced all her contacts. She attends a tutorial school, so the teachers and other kids.
And now they are conducting testing for 860 families who live in her housing block or the grand-daughter’s, or work in the market or work in or attend the tutorial center. At least 5000 people from 1 case!
They actually do something similar whenever we have a local case of dengue fever (not endemic here) – so they do have practice.
Jerri-Lynn Scofield: So, that’s what test and trace means! And not via an app.
Sarah Borwein: No, not via an App
Although there are websites where you can see the locations of all the positive cases, and any flights etc (including seat number) they have been on – so you can self-report if you were near them. But mainly they do the shoe-leather work as the mainstay.
I’m afraid that just turning this process over to an app that notes when a person was in the presence of someone who tested positive is not the same thing. Charavarti discussed problems of trust and privacy in three cases, South Korea, India, and Germany, but doesn’t seem to understand there is a flaw in the contact tracing by app approach.
South Korea’s success involves combining the tracking app and correlating it with lots of additional surveillance data. I have not studied their system, but understand that surveillance information is regularly collected about people, and the app that tells when you were in the presence someone who tested positive is only one part of the country’s contact tracing approach. And that’s not the only thing the country has done. Not only did South Korea test widely, but as Chakravorti acknowledges:
Its contact tracing arsenal included tracking apps paired with CCTV footage, travel and medical records and credit card transaction information.
India, for its part, has virtually abandoned the effort as there are considerable concerns about safeguarding the data that is collected. But there are lingering legacy aspects. Originally it was a requirement that one download and use the app before boarding a ‘plane. International flights remain suspended, but there’s confusion about whether the app requirement still stands, with some immigration officials still insisting it does apply. So the prudent person who wants to fly out of India when that again becomes possible will preemptively download the app.
As for Germany, with its unfortunate history of misuse of data first by the Nazis and then by the Stasi, its citizens are rightly skeptical of contact tracing via app. So, what do they do instead? Seems to be a variation of that old shoe leather epidemiology also practiced by Hong Kong and discussed by Sarah above, Per The Conversation:
Even with a decentralized, privacy-protecting approach, Germany’s new app is unlikely to achieve the level of adoption of South Korea’s. However, the government’s investment in an effective traditional contact tracing approach using public health staff to investigate contacts makes a digital alternative less urgent.
Contact Tracing Theater: Apps
My resistance to use of apps for contact tracing purposes stems from more then idle unconsidered civil liberties concerns – although privacy and controls on data collection are undoubtedly important and should not be surrendered lightly. But just as the security theater we’ve all come to loathe doesn’t really make you any safer when you eventually board the airplane, so contact tracing by app is not very sound contact tracing at all. It just makes big Tech companies and their executives richer, more powerful, and more central to our existence.
31 comments
Comments are closed.
FYI the post has some double spacing between words within sentences that look unintentional.
Either this was fixed (I didn’t) or this is a local issue, since I don’t see it when I view the post.
@ason Boxman
July 14, 2020 at 2:04 pm
——-
I don’t see it, either. Using the Opera browser on Win 10 Pro.
More to the point, when is Jerry-Lynn going to finish writing her book on textile artisans.
I feel like it’s taking forever. I am sick and tired of covid all the time, where are the Textile Artisans?
Ha! I was going to upload a post about sustainably produced organic masks and the textile artisans who make them, but I bumped it for more pressing covid stuff.
And a confession: I’ve shelved the textile book for the time being, and am just about finished with a mystery novel, Death in the Deep Red Sea. But glad to see someone is still interested in that textile book. Back to that soon! I’ve a commission to edit one month of the Eastern Review devoted to Indian textiles. Later this year. Will serve as a dry run for part of the book.
At least until then, I’ll stay focused on covid, among other things (e.g., plastics, right to repair, sustainability, neo-liberal depredations, the legal and regulatory news du Jour, whatever else piques my interest. etc.).
Once again, Jerri-Lynn, you have taken down a technofix fairy, just like you have with recycling.
We have, in the US, at least, become indoctrinated with the idea that Big Tech has the answers and can always make our lives better. Not true.
Many times the important questions have been answered by decades or more of experience “on the ground”, and, as in this case, good ole shoe leather and lots of helpers leads to the best results.
One case generates 8000 contacts? I think you can see that with 15,000 cases/day in FLA, that is clearly impossible. You might as well just assume everyone there has the darned disease.
And that means a lockdown. Or crash the hospitals. Looks like they have gone for Door Number 2.
The number of cases that generate thousands of contacts is small – most people have a small number of contacts or none, and don’t infect anybody else. The fact that the majority of transmission is down to a small number of people is what leads to the ‘cluster’ dynamic.
However, you’re right that it’s not the answer when you have a problem on the scale of Florida, at least not on its own. When the virus has spread that much, modifying the social fabric to reduce transmission is your only real option – be it some variant of lockdown, masking, social distancing, a combination etc. The stricter you are about this, the quicker you can get it down to a manageable level (New Zealand did it in about 2 months, Florida is starting from a worse spot so I think 4-5 months would be the minimum).
What a solid contact tracing infrastructure can do is answer the question of: “What next?” If you can’t answer that question, people will suspect (rightly) that your lockdown is just a stopgap rather than a permanent solution and compliance will be poor.
To summarize, the playbook is:
1. Enact social measures like lockdown to reduce infections to a low enough level to be manageable
2. Use the time until #1 is completed to put a robust contact tracing infrastructure in place (the goal in New Zealand was all contacts traced within 3 days – they were at 4 days when the lockdown ended, but since it was more successful than anticipated they weren’t challenged right away, and are probably at the benchmark by now)
3. Return to a slightly more constrained version of normal, have a check-in process so people can track places they visit, and have your contact tracing teams ready to jump on any cases of community transmission like a ton of bricks to get them under control quickly
4. If you fail at #3 at any stage, go to #1, otherwise remain at #3 until countermeasures become available
This is how countries like South Korea and Hong Kong are attempting to manage it, mostly successfully (at least far more so than the US). It’s also what New Zealand is attempting to do – the test will be what happens when we get community transmission again, as we almost certainly will.
Currently the US strategy seems to have been an Underpants Gnomes variant of the above (1. Lockdown 2. ??? 3. Profit!) which I think is a big part of what’s driving the reopening and resistance to lockdown measures.
In my humble opinion it is too late for contact tracing to work in the United States. The virus has spread too far to make any approach viable. The virus will not be contained until either herd immunity is achieved or a safe and effective vaccine is widely available. In the meantime, we must continue mask wearing and social distancing in order to buy time for a vaccine to be available.
Thanks for the article.
Just a few thoughts: the app presumably does not know whether the contacts were masked or not. Still relevant to trace all contacts, but might give a prioritisation.
Also I gather the German app has been quite successful to date given the historical caveats, but it is also supposedly carefully constructed to not centralise data.
Many people live in fairly compact social graphs. 5,000 traces for one case might still be 5,000 traces for 50 cases. The problem of superspreaders is both transmitting to many at once and, worse, transmitting to varied groups. A mega-church where many smaller graphs intersect or cabin crew / healthcare workers who meet many members of the public from all over a region or internationally spread are a tracing nightmare. Cab drivers might be worst – confined spaces, mobility of driver and passengers etc.
Most US states are far short of the number of contract tracing personnel needed to keep up with current positive test numbers, but even if we had the necessary tracers no would answer their calls due to the robocall plague unleashed on US consumers by profit hungry carriers.
in the olde days, we used to leave extensive messages relaying who we were, why we were calling, and what we were in need of knowing if the receiver chose to call us back.
but few people, in representing an organization, corporate or civic, seem to understand how to leave a message anymore. and even fewer know how to listen to their own messages and glean these important details before deciding whether to return a call. or no one wants to take the time, which ends up being the same thing.
Yet another unintended consequence: the crapification of private phone lines as a communication channel due to rampant telemarketing. I never answered my phone when I lived in the US either.
(And yes, I have seen this raised as a concern in articles about contact tracing attempts in the US).
Used to do some contact tracing, which was routine.
I had a paragraph from the Indiana State Regs handy which clearly stated that urgent matters of public health superceded all other confidentiality laws.
So I could call their doctor. Back then more people had doctors, and answered their phones occasionally.
If I was stretched too thin or there was an outbreak of something the epidemiologists would come down from the state to help. This was 2004-2006. Southern Indiana. And they would borrow from federal. The Acting head of the state Dept of health was borrowed from the Fed at that time.
Wonder how this would have worked in the 70s and 80s with VD.
“In fact, contact tracing is not simple”
and that is a HUGE understatement!
Add in to the mix severe stigma around the diagnosis of COVID (or other infectious diseases like TB) that exist in some communities and you have a very challeging job. Not a job likely to be well done by some 18 year old with a three week online course under his or her belt. It requires nuance, patience, kindness, firmness, and ideally a deep knowledge about the community one is in.
People not only can often NOT remember what they did last week and who they met with, they DONT WANT TO remember
+100
A granular story about working as a nurse in public health. Keep in mind that these were the good times, relatively speaking. Contact tracing, testing, treating were all intertwined.
In a two year period I only had about 12 active TB cases. Here’s one. This will give an idea of how complex public health is.
And fragile.
This info is as of 2006. While I am sure that there have been advances in treatment, I am equally sure that they have not been implemented because $. (Lazy, I know. Prove me wrong.)
My fear is that, when the coronavirus tide finally goes out we’ll discover outbreaks of other contagious diseases
that we’d previously had somewhat of a handle on, TB among them.
TB Control
First, a bit about the illness. Kind of hard to catch compared to SARS-COV-2. Takes 2 hours in a poorly ventilated room with a symptomatic person with active illness.
Second, very hard to get rid of. Can be Latent (hidden, asymptomatic) for years, then burst forth when the immune system declines. Antibiotic resistance is an issue. Usually treated by a cocktail of antibiotics and usually for six months.
Third, it is estimated that 20% of the world’s population has it, mostly in its latent form.
One of my patients, from Mexico City, had severe illness in all four lobes. He was in Home Isolation. His father had died in Mexico “of a coughing illness”,- his mother was ill with, “a coughing illness.” He had come, with his FIVE brothers, to try to make it in the US. They rented a house, kept the AC on 24/7 and the windows closed. I went in with my N95 mask on and daily watched him swallow the 13 pills that would cure him. He was actively contagious. (Positive sputum specimens.)
One time, when we were sitting there quietly, after he took his pills, I started to hear snoring. It wasn’t coming from either one of us. His brothers were all gone to work on their bicycles, mostly to restaurants. So who was snoring? I leaned to the side in my chair and saw feet behind the sofa my client was sitting on. I asked him- oh this was a friend who’d just come to town last night from Mexico City.
This meant I needed to do a TB skin test on the snoring guy. I would go out to the car and get my little refrigerated bag and administer it, but then I had to find snoring guy again after 48 hours (but before 72 hours) to check it.
I already knew it was going to be positive because Mexico’s public health protocol is to give the TB (BCG)
“vaccine” which guarantees a positive skin test thereafter. What I REALLY needed to order was a chest X-ray, but protocol was I had to have a positive skin test result first.
So I TB skin-tested the snoring guy
just as I had tested all the brothers, a brother’s girlfriend, the girlfriend’s daughter, etc.
Things got slippery when I couldn’t find someone at the 48-72 hour window for TB skin test reading and then needed to readminister. Or when the TB skin test was positive and I had to get them to the hospital for an Xray which would finally tell me if there was a suggestion of active disease. My new client might be hard to find. Even then they were afraid.
The building that the Health Dept Clinic was housed in had a collection of “revenue neutral” programs,-
HIV Testing, childhood immunizations, infant car seat program, etc.
In the Latino Clinic down the hall were ladies who were there for Pre-Natal care. Rapport was such that the staff there were pretty good at figuring out who was related to who. Nowhere in this process was the issue of being in the US “sin papeles” (without papers) more than mentioned in passing. TB was the enemy.
I would ask the staff there to seek a possible family member who could help find my client. Next thing I would know, “Verónica’s nephew” would show up at my door for his chest X-ray. I would drive him to the hospital and get it done.
One of my rapport moves with these clients was to share with them that my great grandparents’ generation had come over from Ireland in “Steerage”,-
inside the hold of the ship. Within two years of arrival 4 members of that generation had died of “consumption”
(TB). My father had TB, extrapulmonary, on a kidney, contracted most likely in childhood. It was discovered and successfully treated when he joined the Navy for WWII.
Disclosing this helped my relationship
with my clients.
My guy took about 2 months to have a “clear” sputum specimen come back so that he could return to work. That just meant he wasn’t infectious, but he was not yet cured. I would usually meet him at home in the morning to observe him taking his pills. Once or twice I met him behind the place he worked. None of the other brothers had active illness though one or two had the latent, non contagious form. Treatment for that, though strongly encouraged, is voluntary. They chose not to.
The United States fell flat on its face and 140,000 Americans have died so far. The reasons; first, government was dismantled. Then science dismissed. Messaging politicized. Finally, testing is a debacle.
To regain control of the pandemic, testing must be sped up. Antigen tests for virus proteins similar to a pregnancy test which can done in minutes are on the horizon which can end the lockdowns. They will allow testing at work and school. Those who test negative can be isolated into clean communities. If positive, the longer tests for the virus itself must be conducted and the person isolated. Two negative tests and one is free to return to school, work and community. If they remain positive, the person must be quarantined until medically assured that they are not shedding the virus.
Those who are positive must have their contacts traced. No only by phone but in the field. Data collected must be centralized to assure it is factual and searchable. It cannot be used to make a profit.
To remain one nation, the USA must centralize the fight against the coronavirus pandemic. Congress must appropriate the money to make this possible. A President who understands science and can appoint up-to-speed staff to administer it, is necessary too.
If not, the succession of the former United States into clean and infected states is inevitable.
I was under the impression the antigens after COVID are short lived and disappear after just a few months.
This was mentioned again on last night’s NBC news. The low levels and short duration of antigens indicate that reinfection is a possibility and a lifetime vaccine unlikely. Perhaps medical workers and high-risk workers will get shots semi-annually or more frequently. The mostly likely workable scenario is weekly antibody tests at work and school with contact tracing of the positive results, isolation and health clinics to keep track of everyone and provide treatment once one is found.
Even if a treatment is discovered, a federal public health system will be needed to keep the population healthy and pay for the treatment. Coronavirus is a highly contagious common cold that is roughly 10 times more deadly than seasonal flu or HIV in the USA. Think of the public health hospitals, quarantine locations, systems and programs needed to avoid a sick, isolated and reactionary society in the years ahead.
Every network station has their own medical contributors. The medical contributor for CBS – Dr. David Agus is very optimistic that a vaccine will be available, and that it will provide long term immunity. He mentioned the importance of T-cells with memory in fighting the coronavirus. Antibodies are only one part of our immune system, and not even the most important part.
No, you are thinking of antibodies. Antigens refers generically to any protein shed by the infection, not necessarily antibodies.
I know of someone near D.C. who tested positive, but was asymptomatic, and reportedly did not alter behavior. I think this must be a serious complication for test and trace plans in the U.S.
There is also another kind of tracing which is to test the sewage for the disease. At this point so many Americans are infected, I wonder if we are past the point of no return on the tracing strategy. We don’t know how many Americans are infected. It may be as high as 20 million or more, or around 1%, although probably not uniformly spread out.
I meant 10%, not 1%.
Wouldn’t that be more like 20/330 = .06?
And that 20 million would be the number of people who have caught the disease to date. Those who have recovered or died, and are thus no longer relevant to further spread don’t count.
What you need here is an estimate of the number of currently infectious cases who are not in controlled circumstances (eg. hospital, quarantine) for your numerator.
That should reduce the number of active cases to be traced substantially.
Still, you are right that the large number of cases is a measure of the problem of tracing.
On the other hand, the US is a large wealthy, largely literate country with a lot of people, It should therefore be capable of fielding a lot of trace teams.
The better measure of the difficulty would be the new infection rate compared to the size of the population to provide trace workers and the technical and financial resources to support the tracing effort.
The 20 million figure (if it is correct) is dated, so I assume the numbers have increased, hence “around 10%”. The idea behind contact tracing, I assume, is to isolate infected people. At a certain point, when too many people are infected, is this still a good strategy? How much COVID immunity is acquired, or the range of possibilities, in people who have recovered, is an open question.
Ok. I have repaired my error (not checking the data to clarify the situation).
As of 1800 today (2020-07-15) the Johns Hopkins coronavirus tracker is reporting about 13.4 million confirmed cases worldwide, and about 3.5 million in the US.
—
CAVEAT: As has been discussed here a number of times, determining the actual number of cases is to some extent impossilble, and numbers from different jurisdictions reporting by local definitions are not comparable. Recurring news articles show numerous errors discovered in estimates of the incidence of disease, ranging from 10 times too high to tens of times – or more – too low.
—
While looking for data on such things I did come across good sources for some fairly good data, including epidemiological data updated every day by 1130, and reflecting the data held at the provincial ministry of health as of 1600 the day before.
I have just begun to assimilate and examine this – the nice people at public health are providing daily and weekly summaries (30 pages of analysis for the latest) and a data tool to do your own data diving, along with downloadable datasets in a couple of formats. available via the provincial data catalogue.
The incidence by age is interesting – between 20 years old and 79 years old every ten year set comes in between 230 to 290 per 100K population cumulative to date. Under 20 the rate plummets, over 79 it soars.
Deaths, however, are significantly more age dependent, with over 49, over 59, over 69 and over 79 ten year age groups each seeing substantially more deaths than the previous group, as a percentage of confirmed cases.
In related news, at an Ottawa city council meeting today, Ottawa’s medical officer of health is warning residents to be prepared to live with the risk of COVID-19 spreading in the community well into 2021 or even 2022.
So far, closing non-essential businesses, working from home, mandating masks in areas accessible to the public, instituting new health safety rules, installing barriers and partition, mandating a 2m separation, and related measures have been generally effective, with the rate of new cases down to a bit more than 10% of the peak
The number hospitalized for Covid-19 is down to 115 today, about 11% of the peak value.
There seems to be a general consensus that special measures will have to persist into 2021 or 2022.
In related news 29 members of Congress in the US have advocated a program for re-opening the border for non-essential travel. People up here are not really in favour of that… not surprising as the confirmed rate in the US is three times higher than here, and we see a lot of news about people in the US demanding the right to act dangerously. Given that about 80% of us want the border to stay closed I suspect government will be reluctant to change that right now.
Still, having worked from home since March, I am wondering what another year or more doing so will feel like.
Interesting times.
The 20 million estimate comes from the CDC. They estimate that only about 10% of American cases have been detected. My point is that if there really have been this many American infections, then we may need to reconsider what contact tracing can accomplish.
https://www.npr.org/sections/coronavirus-live-updates/2020/06/25/883520249/cdc-at-least-20-million-americans-have-had-coronavirus-heres-who-s-at-highest-ri
There are a number of critical flaws in automated ‘contact tracing’ projects..
Many of these involve smartphones. So far these seem to use GPS, bluetooth, or wifi.
From what I have read, the chance of infection drops by one half for each additional metre of separation at least as far as 3 metres, possibly more. Unfortunately, many GPS receivers don’t get better than 3 metres accuracy at best, and this is subject to occasional losses of accuracy. Thus, in the best case, the separation of two cell phones is known +/- 6 metres. Thus a reported separation of 7 metres may actually be 1 metre – equivalent to two people holding hands while grasping a cell phone in the other hand. Conversely, a reported separation of 1 metre may easily be 7 metres in the real world.
I’m one of the techno-curious types who occasionally puts GPS systems into a mode that shows the accuracy of the locaton at the current moment. In many cases, that accuracy bounces about between 3 and 7 metres for a good GPS in a location giving a good view of the sky, and thus direct paths to a number of satellites, at a good signal strength. In cities, and certain types of terrain, there can be an issue with signals bouncing off one feature while the direct path is blocked by another feature (terrain, buildings, parts of vehicles, etc.). Clearly the level of uncertainty here makes it next to impossible to avoid a myriad of false negatives and false positives.
Systems trying to estimate distance on the basis of the strength of bluetooth signals are exposed to a variety of factors that may cause a bad guess at the distance, with similar results.
And as others have noted, other circumstances are not captured. Are people masked? Using other PPE? Are there barriers? Is there a good or bad pattern of air circulation? Is the phone in a hand or pocket, in a purse on a nearby table, beside your computer or front door? Are the people 2 metres apart singing for an hour, or has someone infected coughed in the area in the last shortish while, contaminating surfaces, and possibly the air for a sufficient period?
And, of course, what data is collected, what groups get it, how long is it kept, and how will/can it be misused?
I don’t see a lot of benefit, but I can see a lot of various risks.